|
|
January 2022 Issue Vol.12 No.1
SELECTION OF CLASSIFIER MODELS FOR INTRUSION DETECTION SYSTEM (IDS)
https://ia601500.us.archive.org/0/items/vol12no101/vol12no101.pdf
Mrs.V.Mounika 1,Research scholar Department of Computer Science and Engineering,
Koneru Lakshmaiah Education Foundation, Vaddeswaram, AP, India
Dr.N.Raghavendra Sai 1,Assoc.Professor Department of Computer Science and Engineering,
Koneru Lakshmaiah Education Foundation, Vaddeswaram, AP, India
Abstract:
Any unusual move can be considered a break in quirks. Some procedures and calculations were mentioned in the drafting to identify irregularities. In most cases, true positive and false positive limits were used to observe their display. However, depending on the application, an off-base false positive or false positive can have serious adverse repercussions. This requires the incorporation of cost-sensitive limits on display. Furthermore, the more popular KDD-CUP-99 test data set has a huge information size that requires some pre-management measure. Our work in this article begins by listing the need for a delicate cost examination with some original models. After talking about the KDD-CUP-99, a methodology for the end of the reflections is proposed and later the possibility of reducing the amount of the most significant reflections in a simple way and the size of the KDD-CUP-99 in a indirect way. From the revealed writing, the general techniques are chosen to detect the irregularities that best behave for the various types of aggressions. These various filing cabinets are stacked to frame a team. An expensive method is proposed to dispense the relative loads to the classifiers equipped for the realization of the finished product. The profitability of the false and genuine positive results is performed and a technique is proposed to choose the components of the profitability measures to further improve the results and achieve the best overall exposure. There is talk of the effect on the exchange of execution due to the merger of the viability of the expense.
Keywords: True positive (TP), False Positive(FP), Support Vector Machine (SVM).Intrusion detection system (IDS)
